Megan M. Brown-Bousfield

CLI AND TRANSFER IN A TRILINGUAL CONTEXT: ACQUISITION AND DEVELOPMENT OF L3 GERMAN GRAMMATICAL GENDER

MEGAN MARISA BROWN-BOUSFIELD

Boston University, Graduate School of Arts & Sciences, 2024
Major Professors: Paul Hagstrom, PhD Associate Professor of Linguistics
Charles B. Chang, PhD Associate Professor of Linguistics
Suzanne Flynn, PhD Professor of Linguistics and Language Acquisition, Massachusetts Institute of Technology

ABSTRACT

In this thesis, I compared two groups of adult beginner German learners acquiring German in a college or university class context. One group of participants consisted of English/Romance bilinguals learning German as an L3 (the “L3 German” group), where, of the learners’ two previously known languages, the Romance language has a grammatical gender system. However, the more typologically similar language, English, has no such gender system. A second group of participants consisted of a control group of monolingual English speakers learning German as their L2 (the “L2 German” group), all of whom have no previous experience with a gendered language.

Across two experimental sessions, participants completed three distinct experimental tasks designed to measure participants’ sensitivity to German gender errors: a gender assignment task, a forced-choice listening task, and a self-paced reading task. Overall participants across both language background groups were ultimately able to develop at least some sensitivity to the German grammatical gender errors presented in the experimental tasks. However, participants in the L3 German group generally demonstrated more evidence of sensitivity to gender errors than L2 German participants, suggesting some degree of CLI or transfer of previously acquired grammatical gender knowledge from those learners’ Romance language to their new L3 German gender system. Additionally, within the L3 German participant group, the age at which participants acquired their Romance language also impacted German
gender sensitivity. Participants who acquired their Romance language at a younger age demonstrated higher sensitivity to German gender errors than participants who acquired their Romance language at an older age.

These findings have implications for models of L3 acquisition and development, some of which propose that the order in which previously known languages were acquired impacts the degree to which each language may influence L3 grammatical development (e.g. Bardel & Falk, 2007, et seq.). This author proposes the breaking down of the concept of “order of acquisition” of previously acquired languages in the L3 literature into more precise variables that can be measured and compared in a more precise manner, for example by instead examining acquisition and language
proficiency as was done in this dissertation.